What is the Sandbox?

This "Sandbox" is a place where Code Golf users can get feedback on prospective challenges they wish to post to the main page. This is useful because writing a clear and fully specified challenge on the first try can be difficult. There is a much better chance of your challenge being well received if you post it in the Sandbox first.

To post to the Sandbox, scroll to the bottom of this page or click on the "Add Proposal" link below, and click "Answer This Question". Click "OK" when it asks if you really want to add another answer. Write your challenge just as you would when actually posting it. You may also add some notes about specific things you would like to clarify before posting it. Other users will help you improve your challenge by rating and discussing it. When you think your challenge is ready for the public, go ahead and post it, replace the post here with a link to the challenge and delete it.

See the Sandbox FAQ for more information on how to use the Sandbox.

Get the Sandbox Viewer to view the sandbox more easily

To add an inline tag to a proposal use shortcut link syntax with a prefix: [tag:king-of-the-hill]

Invert the Fundamental Transformation on permutations code-golf

A permutation like

[3, 5, 1, 2, 4, 6]
1->3, 2->5, 3->1, 4->2, 5->4, 6->6


can be decomposed into cycles

(3, 1) (5, 4, 2) (6)


There's many ways to write these cycles, but we can fix a canonical form by requiring that:

• In each cycle, the greatest element is written first.
• The cycles are sorted in increasing order of greatest element.

In this form, we don't even need parentheses to demarcate the cycles. Just from the list of numbers

3, 1, 5, 4, 2, 6


we can infer that the cycles starts at the numbers that are greater than any to their left.

 3, 1, 5, 4, 2, 6
*     *        *
(3, 1)(5, 4, 2)(6)


From here, the original permutation [3, 5, 1, 2, 4, 6] can be uniquely recovered, and your task is to write code that performs this recovery.

Input: A list of n distinct numbers 1 through n. Optionally, the value n itself.

Output: The permutation f that produces that canonical form cycle list, listed or printed as [f(1), f(2), ..., f(n)]. Format is up to you.

You may optionally work with 0-indexed values 0 ... n-1 instead of 1 ... n.

3-Hole Golf Course: Text Formatter

This is a series of three code golf challenges. The over-arching idea is that in each of them, you write a program that processes text in some way, and if you pipe an input text through all three solutions (with a small amount of glue code), it spits out a nicely formatted paragraph.

Hole 2: Line Wrap

You are given as input a hyphenated piece of text, and the maximum width n ≥ 2 of a line. The output shall be a newline-delimited string, obtained by greedily splitting the input into lines of length at most n characters. Lines can be broken between words (replacing a space with a newline), after a tilde (replacing it with a hyphen), or after a hyphen (inserting a newline after it; this is needed with words like well-known). All tildes which are not replaced by hyphens should be removed, and this should be taken into account when computing the lengths of the lines. The correct output for Hy~phen~ate this sen~tence! and 20 would be

Hyphenate this sen-
tence!


A syllable that's at least n characters long and not the last syllable of a word shall be cut after the n-1st character, and a hyphen shall be inserted there before the line break. The same applies to the last syllable of a word that's over n characters long.

Hole 3: Justification

You are given as input one line of text, and the desired width n of a line, which is at least the length of the input. Your output is that line, but with extra spaces added between the words to make its length exactly n. The extra space should be added in an aesthetically pleasing way, so we require the following.

• The lengths of any two runs of spaces differ by at most 1. This means that for some number k, every run of spaces between two words shall be of length k or k+1. For example, we could have runs of only 5 or 6 spaces between the words.
• The list of lengths of these runs shall be balanced, in the following sense. For any two contiguous sublists of the same length, the number of ks in them differs by at most 1. For example, the list of lengths 5, 6, 5, 5, 6, 6, 5 is not balanced, since the length-2 sublist 5, 5 and 6, 6 contain 2 and 0 occurrences of 5, respectively. The list 5, 6, 5, 5, 6, 5, 6, on the other hand, is balanced.

These constraints do not uniquely specify the output. For example, for the inputs Hyphenate this sen- and 20, the two acceptable outputs are

Hyphenate  this sen-


and

Hyphenate this  sen-


If the input contains only one word, it should be left unchanged.

Sandbox notes

In all of the above holes, we can add some constraints to the input, if the challenge would otherwise be too complex (all hyphens and spaces occur between letters, all punctuation occurs at the end of words etc). However, I'm a little concerned that the holes may still be too complex. In the first hole, I can relax the case-insensitivity of the hyphenation list, and in the last hole, the second requirement can be dropped, but I feel that the second hole is already about as simple as I can make it.

• I'm confused about the first hole: if hyphens are a known problem, why does the hyphenated word input use hyphens rather than tildes? In the second hole, for consistency I think you should replace "at a tilde" with "after a tilde". The cut at the n-1st character doesn't seem complex to me: it's just a case of how the "where should the cut be?" variable is initialised. – Peter Taylor Aug 25 '15 at 20:19
• Also, I think the given definition of balanced in the third hole doesn't say what it's intended to say. In particular, subsequence is almost certainly not the word you want to use. – Peter Taylor Aug 25 '15 at 20:26
• @PeterTaylor Thanks for the input! I tried to clarify the balance condition; I think it says what I want it to say, but it's hard to write clearly. – Zgarb Aug 25 '15 at 21:17
• The point is that 2, 2 is a subsequence of 1, 1, 2, 1, 1, 2. Perhaps you could use the word sublist? – Peter Taylor Aug 26 '15 at 5:19
• @PeterTaylor Oh, I see what you mean. Edited. – Zgarb Aug 26 '15 at 14:04

Enough jQuery?

As a general rule of thumb, jQuery is a vital part of any program. It is really great and does all things. Your job, as a quality-control expert, is to ensure that a given program meets industry standards. Only programs that use jQuery are eligible for your seal of approval.

The Task

Write a program which takes an HTML file as input, and returns a High-Quality (TM) HTML file as output. An existing HTML file is deemed as high-quality if it contains a script tag with the word jquery in it. Examples include the following:

<script src="jquery.js"></script>
<script src="//code.jquery.com/jquery-1.11.1.min.js"></script>
<script src="https://ajax.googleapis.com/ajax/libs/jquery/2.1.1/jquery.min.js"></script>


This condition is satisfied whenever there is the string <script src=", followed by "></script>, with the string jquery located somewhere in the middle.

(I'm not sure if I should make this condition more precise. It's probably good enough.)

If the given file meets quality standards, simply output it as-is. If there is no enough jQuery, however, output the file with the string <script src="jquery.js"></script> at the very top, so that everyone can see that it's a fixed, bug-free program.

• <script type="text/javascript" src="..."></script>? – Peter Taylor Sep 6 '15 at 19:42

Blastoff

Your goal is the create the following text output:

10 9 8 7 6 5 4 3 2 1 Blastoff!


The "10 " will be generated by the following C program:

#include <stdio.h>
int main(){printf("10 ");}


The "9 " will then be generated by a function in a different language of your choice; the "8 " by a function in a different language; and so on until the last function generates "Blastoff!". Each function will take as input the complete source code of the previous step in the form of a string and output the next number in the sequence and a space (or the string "Blastoff!").

Each function must "do something meaningful" with the input. Doing something meaningful is defined to mean that the output will vary based on the input. That is, ignoring the input or trivially using it (e.g. making it evaluate as 0 no matter what comes in and then adding the needed number) fails to meet the challenge.

Scoring

A contestants base score will be the sub-total of the byte-length of all 10 functions. To encourage creativity, the base score will then be divided by the number of different techniques used to generate the output. For example, "string splitting", "character count", "bitwise and", etc. The technique of each function is defined as first thing done to the input string other than assignment to a variable.

Example

The following Perl code illustrates how a "9 " can be generated:

nine('#include <stdio.h>
int main(){printf("10 ");}');

sub nine{
@a=split(/"/,$_[0]); print$a[1]-1 ." ";
}


Here, the function nine() takes the source code of the C program and manipulates it to get the number 9, then adds a space to the printed output. For scoring purposes, only the subroutine counts - the additional code was added for illustration purposes only and need not be listed on an entry in general. The input string for the next function will start with the "s" in "sub" and go through the final "}"

The score for this function is thus 54 (including the actually unnecessary CRs and spaces). The technique used here is string splitting. An entry that included this code would thus look like this:

9: Perl

sub nine{
@a=split(/"/,$_[0]); print$a[1]-1 ." ";
}


Score: 54

Technique: String split

...

Total

Base score: (54 + ... ) = 512

Unique techniques: 8

Final score: (512 / 8) = 64

Meta comments:

• The goal of the rules is to 1) avoid trivial solutions and 2) encourage thinking about the next step while writing the current one. If the rules need to be added to enhance either aim, let me know.
• Is ten functions too many? I could start the count at 5 if that seems better
• I think 10 is good, as it encourages users to be ever more creative with each function. – mbomb007 Sep 10 '15 at 20:44
• "Doing something meaningful" is very hard to make precise. – Peter Taylor Sep 10 '15 at 20:54
• @Peter I defined it as having variable output based on the input. While something as simple as an if statement will cause that to happen, I think it might be OK when combined with the incentive for different techniques - 10 trivial "if input >=< something" routines won't be a very good score anyway as all techniques will be "comparison". Of course I'm open to better wording suggestions. :) – ThaddeusB Sep 10 '15 at 21:29
• @ThaddeusB Great idea. +1 – mınxomaτ Sep 16 '15 at 3:21

The Algebra of Reflecting Points

This is a challenge based on manipulating points with a specific set of operations, each dealing with the reflection of some points over others.

Warning: There's not actually a challenge here yet, just the basis for a challenge that could be to "simplify the given expression" or something.

(I'll add more diagrams later)

Lists of Points

The fundamental object is an ordered list of points, like (A,B,C) or ([2,3],[5,8],[6,8]).

Reflection

A r B represents the reflection of point A across point B, resulting in a new point C so that B is the midpoint of AC.

A r (B,C,D) represents the reflection over a series of points, and is equivalent to A r B r C r D.

(A,B,C) r (D,E) represents (Ar(D,E),Br(D,E),Cr(D,E)), with either list being of any positive length.

The result of the reflection operator is a list of points that is the same length as the first operand. (If the first operand is a single point, then the output is a single point.)

A list with a single point is that same as that single point. (B) == B

Lists can be arbitrary grouped inside of other lists. (A,B,C,D) == (A,(B,C),D) == ((A,B,C),D)

Simplification

A point reflected over itself is an identity, CrC is the same as C. Any point reflected over the same point twice is an identity, rCrC can be removed.

(A,B,C) equals (C,B,A)

For any three points ABC there exists a unique fourth point D=C-B+A such that anything r (A,B,C) = anything r D. This means that any long chain of ArBrCrDrErF... can be reduced to have fewer than three rs.

((A,...,B)r(C,...,D)) == ((D,...,C),(A,...,B),(C,...,D)) and (ArB) == (B,A,B)

Examples

Show that Cr(CrA)rB == Cr(BrArC)

Cr(CrA)rB         #original
Cr(A,C,A)rB       #expanding 4th simplification rule
CrArCrArB         #the list is equivalent to a series of rs
CrAr(C,A,B)       #grouping to form a list
CrAr(B,A,C)       #swapping, the 2nd simplification rule
CrArBrArC         #expanding list
CrCrArBrArC       #Identity operation of C=CrC
Cr(C,A,B,A,C)     #listifying
Cr((C,A),B,(A,C)) #further grouping
Cr(Br(A,C))       #using the 4th simplification rule
Cr(BrArC)         #expanding parenthesis


Here is this proof visualized geometrically.

Whose Llama is it anyways?

BetaDecay posted a legit looking movie poster in chat, which got me into thinking that this could be a very nice challenge!

Your task is to overlay a nifty llama poster on top of a movie poster in a way that it still looks a legitimate movie poster. The image posted by Beta Decay is:

a http://pictures.boxxspring.com/pictures/960x0/100588

Your program will be provided with an input of fixed size (TBD) movie poster of any popular movie and a fixed llama cutout to overlay that poster as arguments. You may scale (proportionally), rotate or translate the llama cutout anywhere on the movie poster to make it look like a llama is photobombing the poster. At the same time, the output image should still look like a legitimate movie poster in a way the above poster feels real. You cannot perform any operations on the movie poster and no other operations on the cutout other than scaling, rotation and translation.

This is a popularity contest, so the answer with the most net votes wins. Voters are encouraged to judge answers by:

• The correct placement of the cutout such that it does not outright look like a cutout
• The scaling of the cutout to match with people/objects in the movie poster
• The placement of the cutout with respect to the movie text. i.e. The cutout should not hide the movie title in a way that its no longer understandable.

Input

Two images in any common image format. The input can come as paths to the images or the images themselves (if your language support image input) either as function arguments, ARGV, STDIN or equivalent.

The first image (movie poster) will be of fixed resolution and the second image (the llama cutout) will be of enough resolution in order to have good quality even after scaling or rotation.

Output

A single image of the same resolution as the movie poster image in any favorable image format.

Llama Cutout

Please use this cutout as the second argument in your program:

[TBD]

Test Images

Here are a few images to test your algorithm on. Click on the images to get the actual resolution.

TBD

Please include some examples for these test images (along with other examples if needed) in your submission. Its also a good practice to briefly explain your algorithm as well.

Help Agent Zigzag with his secret messages

The WW2 double agent Eddie Chapman — aka "Agent Zigzag" — used an unusual cipher to communicate with his German controllers. It is described in the appendix of Ben Macintyre's book about Chapman and can also be viewed at the National Archives, apparently.

Read the following description and write a program to encrypt and decrypt messages using this system.

Your program should accept three inputs via stdin and/or command line switches:

1. A keyword
2. A flag to select either encryption or decryption
3. A message to be encrypted or decrypted

The output should be the corresponding ciphertext/plaintext, arranged into groups of 5 letters separated by single spaces. You may assume that the keyword consists entirely of uppercase letters, but should accept plaintexts and ciphertexts containing mixed case characters, punctuation and spaces. Non-alphabet characters should be stripped from the input and should not appear in the output (which may only contain groups of 5 uppercase characters separated by spaces, with an optional line break at the end). The date value used for keyword processing should be obtained from the system clock using the local time zone.

This is a code-golf challenge. The shortest code (measured in bytes) will win.

1. Keyword processing

Write out the keyword, and below each letter write the position in which this letter would appear if the letters were all sorted alphabetically. If the same letter appears more than once, number them from left to right. For example, the 14-letter keyword CONSTANTINOPLE would be processed as follows:

C   O   N   S   T   A   N   T   I   N   O   P   L   E
2   9   6   12  13  1   7   14  4   8   10  11  5   3


(Notice how the three Ns are numbered 6, 7 and 8.) Combine this sequence of digits into a single number, and multiply by the current date (from 1 to 31). For example, if the transmission is being made on the 8th of the month:

2961213171448101153 x 8 = 23689705371584809224


2. Encryption

Pad the secret message to a multiple of five characters in length by appending the null character X as many times as necessary, then encrypt Vigenere-style by cycling each letter through the alphabet by offsets corresponding to successive digits of the number calculated at step 1. For example, the message

Have arrived safely and in good health. Awaiting further instructions.


is encrypted as follows:

havearrivedsafelyandingoodhealthawaitingfurtherinstructionsx << Plaintext
236897053715848092242368970537158480922423689705371584809224 << Key (repeated)
JDBMJYRNYLEXIJMLHCPHKQMWXKHJDSUMIAIICKPKHXXBQLRNQZUWCGBIXPUB << Ciphertext


because 'H'+2='J', 'A'+3='D', 'V'+6='B', etc.

Finally, read off the encrypted message in groups of five characters:

JDBMJ YRNYL EXIJM LHCPH KQMWX KHJDS UMIAI ICKPK HXXBQ LRNQZ UWCGB IXPUB


• For languages that don't have system clock access, is it acceptable to take the date/time as an additional input? – Geobits Sep 23 '15 at 14:09
• @Geobits Would you mind if I said no? We have so many code-golf questions that are dominated by esolangs, I thought it would make a change to have one that favours higher-level languages. (Although it looks like Pyth can do dates...) – r3mainer Sep 23 '15 at 14:55
• I don't personally mind since I don't tend to use them. Some might, so I figured it was better to ask/clarify while it was sandboxed rather than on main. – Geobits Sep 23 '15 at 14:57
• It's not just about esolangs: removing the direct coupling with the system clock allows test cases which work consistently. – Peter Taylor Sep 23 '15 at 20:19
• @PeterTaylor But only 31 test cases are needed. – r3mainer Sep 23 '15 at 22:02

Well hello there, fellow golfers. The time has come for us to take part in the game. Time for the official Illuminati Confirmation!

Illuminati Confirmed?

Nah, just joking. We actually want to hide even better - by making conspiracy theorists look ridiculous.

For that, you need to create code in whatever language you want to. Your task is to print the Eye of Providence, as an ascii art. The program may take input, but it shouldn't rely on it.

However, there's a catch. You need to hide as many non-obvious references to Illuminati in source code as you can. One particular requirement is that your code must have an "Illuminati confirmed" sentence hidden somewhere(although it doesn't have to be plain string - in fact, it shouldn't; you can have it hidden by placing it as a first letter to every line, etc. - be creative!)

And because it's , the most upvoted and creative code wins!

Now I want feedback before posting it - what else should I have in here? What is not clear? And if you come up with better "lore chit-chat"(I actually had no afflatus) - feel free to help me.

• You need to hide as many non-obvious references to Illuminati in source code as you can: One problem you may run into is that I think it's been proven by the internet at large that you can make anything an Illuminati reference if you twist it enough. – Geobits Sep 24 '15 at 17:22
• Also, pretty sure the Illuminati has already been confirmed since it was a real and historical secret society. – mbomb007 Sep 24 '15 at 17:56
• @mbomb007 Nonsense! – MatthewRock Sep 24 '15 at 18:59
• @Geobits What is your suggestion then? I am honestly counting on community, so some stupid answers like "Well, I have a word 'rock' rock is four characters, and four - 1 = 3, triangle also has 3 sides - Illuminati Confirmed" will be downvoted, and some clever ideas will get attention. – MatthewRock Sep 24 '15 at 19:05
• This is an art / wordgame contest, not a programming contest. I'm not as strongly against this sort of thing as some, but there's a fair chance this will get closed. – Level River St Sep 24 '15 at 21:08
• @steveverrill what I had in mind wasn't "Print the prettiest and funniest image and code, but rather to get people creative. This post is one of my inspirations - Perl and Brainfuck are bottles; C makes heavy use of macros to call the function in a funny manner, and while there are many other "plain" programs, the most creative are the most upvoted. This is similar to mine. There is more of that, buut looks like this shouldn't be closed. – MatthewRock Sep 24 '15 at 22:07
• The 99 bottles of beer question is 4 years old and things were different back then. Nevertheless the issue was raised in the top comment on that question. There was a big attitude change just after the christmas tree challenge, in my opinion probably caused partially by Code Trolling. I'm not saying don't do it, just be aware of the issue. James Webster's recent cake question did quite well, though it did attract some close votes. The recent Back To The Future question was closed immeditately. – Level River St Sep 24 '15 at 22:35
• @steveverrill I'm sorry, I have no idea what you're talking about. Could you please introduce me to the issue of Code Trolling, and provide some examples? – MatthewRock Sep 25 '15 at 8:29
• Code Trolling was a particular class of open ended questions. The tag description is here. codegolf.stackexchange.com/tags/code-trolling/info It got out of hand. The history is too long and emotive to discuss in a comment, but it seems the challenges have been deleted. There are many meta questions on the subject. I've picked this one for you. meta.codegolf.stackexchange.com/q/760/15599 – Level River St Sep 25 '15 at 8:56
• @steveverrill Okay, thanks for your insight. I will think about what you guys said, and maybe post the question - the worst that will happen is that iit will get closed - but hopefully it will spawn some clever and funny answers. – MatthewRock Sep 25 '15 at 9:21
• Will an ASCII art version of the Eye of Providence be added in the question, or will we have to make one ourselves? – ASCIIThenANSI Sep 25 '15 at 15:52
• @ASCIIThenANSI That's what I asked in other topic(well, I did it badly). I'm afraid to lock it to one particular output, because this way they can only get creative with code. What do you think? – MatthewRock Sep 25 '15 at 19:11
• If you designed the exact Eye of Providence that we had to produce, it would go better as codegolf than popularitycontest. A good codegolf would require a fair amount of repetition in the pattern (compressing completely random strings is boring.) I guess it's fair to assume your design would be symmetrical. But since you want a popularity contest, just go for a popularity contest, and see if it flies high or if it gets shot down in flames. There's only one way to find out for sure, and you won't do it with meta posts. That said, Geobits makes a good point, that you need to address. – Level River St Sep 26 '15 at 9:39

1,2,3,4,...n-2,n,n-1

Your goal is, with an input "n", to print the numbers 1-n BUT the last two values are flipped. The numbers must be CSV (Comma separated values)

n must be >= 3 (And an integer)

For example, with an input n="10", your output should be as follows: 1,2,3,4,5,6,7,8,10,9.

An input with n="3", your output should be as follows: 1,3,2.

• Get rid of the requirement that you can't just embed a string, please. How about printing 1,2,3,...,n-2,n,n-1 for a given input n? – Lynn Oct 3 '15 at 11:16
• @Mauris Added this in, thanks for your input :) – Albert Renshaw Oct 3 '15 at 20:52
• Shouldn't it be "last two values" instead of "last two digits"? In the example, the swapped items are 9 and 10, and 10 is not a digit. – Reto Koradi Oct 5 '15 at 5:14
• @RetoKoradi Ahah good catch, edited it thanks! – Albert Renshaw Oct 5 '15 at 7:12
• @FryAmTheEggman Got it in, thanks! – Albert Renshaw Oct 6 '15 at 20:21

Categorical logic [on hold]

Inspired by this question:

Task

The point of this challenge is to create a function or program that takes 3 short phrases and returns the logical assessment of the phrases.

Input

Each of the three phrases will be in the form: [All|Some] [A|B|C] are [A|B|C]., where the two A|B|C elements are different. The string Therefore,{space}, where {space} is a single space character, precedes the third phrase and reduces the initial capital of the third phrase to a lower case letter.

You may choose the input source and the separation of the strings. For example, STDIN with three concatenated phrases, text file, phrases separated by newlines, and taking the phrases as command-line arguments in quotes are all acceptable.

Sample input

With three concatenated phrases as example format.

• Some tees are moos. All moos are yees. Therefore, all tees are yees. (output: falsy)
• All A are B. All A are C. Therefore, some B are C. (output: ????)
• All submissions are answers. Some submissions are winners. Therefore, some answers are winners. (output: truthy)

Output

• Output for a logically correct combination of phrases must be a truthy value.
• Output for a logically incorrect combination must be a falsy value. This is also the case if the input is logically unsound e.g. All A are B. No A are B. ...
• Output for a logically uncertain combination must be the following string: ????

The output may be followed by a single newline.

Clarifications for finicky logic

• Some is a subset of All. The following statement is badly written, but truthy: All A are B. All A are B. Therefore, Some A are B.
• The following statement is also truthy: All A are B. All B are C. Therefore, some C are A.

Scoring

This is , so shortest code in bytes wins.

You may reduce your score by 10% if your values for A|B|C can contain spaces. You may assume that the value for A|B|C will not contain the word {space}are{space}).

You may reduce your score by a further 20% if your function or program also correctly handles No as input as well as All|Some.

Leaderboard

Standard leaderboard code goes here. Tag suggestions are welcome!

• You should add what the expected output is for the sample imputs you give, so that one does not have to find if a sample is logically sound (and possibly be mistaken) – Fatalize Oct 13 '15 at 9:23
• Working on a table of expected output. It's bigger than I had expected, and I haven't yet included 'No' as an option... – pbeentje Oct 13 '15 at 9:42
• What is logically unsound about All A are B. Some A are B.? That's just equivalent to All A are B. There is at least one A. – Peter Taylor Oct 13 '15 at 13:27
• @Peter Taylor, You're right. A couple of prods at the table has turned up logical differences for the same statement in different places. I don't have enough time to fix this right now, so am putting the question on hold until I can make a clear definition of required behaviour for different cases. – pbeentje Oct 13 '15 at 13:58
• "All answers are submissions. Some submissions are winners. Therefore, some answers are winners." This doesn't seem truthy to me. A simple substitution: All cats are animals. Some animals are birds. Therefore, some cats are birds. – Geobits Oct 13 '15 at 17:08
• In the few minutes at a time I'm spending on this, I'm struggling to be logically consistent. I feel that All A are B. Therefore, all B are A is false because of the presence of Therefore in the statement. My gut wants me to categorise All A are B. Therefore, all A are C as logically uncertain, yet applying the same logic this should actually be false (as should all logically uncertain combinations). I'll have a search but would appreciate it if anyone already knows of a definitive source. @Geobits: thank you, I've corrected that example. – pbeentje Oct 14 '15 at 14:08

Two mountain hikers want to meet, but must stay at the same altitude

This question is totally inspired by this question on Puzzling SE. For ease, I have reproduced the original question here:

Two hikers are separated by a two-dimensional mountain range, like the one shown below. The mountain range alternates between peaks and valleys, connected by straight lines.

Both hikers are at sea level, and the mountain range never dips below sea level.

The two hikers want to meet up with each other. Prove that they can do this while staying at the same altitude as each other for their entire journey. They are allowed to backtrack.

Input/Output

Meta: what input/output format would be better?

Continuous:

• Input: a list of ordered pairs (x,y) that represent the peaks and valleys.
• Output: a sequence of positions where one hiker or both changes direction, formatted like so:
(x1 y1) (u1 v1)
(x2 y2) (u2 v2)
(x3 y3) (u3 v3)
...

Discrete:

• Input: a sequence of heights that represent the height of the terrain (with no flat portions). (I.e., 0 1 2 3 4 5 4 3 4 5 6 7 6 5 6 5 4 3 2 3 2 1 0, maybe with steps larger than 1?)
• Output: the left/right moves each hiker has to make, outputted like
L R
L L
L L
R L
...

Scoring

• Scoring is in bytes.
• Bonus: -5% if your program produces optimal solutions.

Meta:

• I'm pretty sure this is not a duplicate, but still, is it?
• I'd like a better title. Would "Mountain Hikers" be good enough?
• Any other feedback?
• Is the input sorted? Also, I don't understand the first output format. – feersum Oct 29 '15 at 16:15
• @feersum: Sorted in the x-direction, yes. The rationale behind the first output format is that as the hikers are moving, they're both moving in one direction (not necessarily together). However, when one reaches a peak or valley, the other will stop and change direction. The output should have the positions of the hikers at these moments. – El'endia Starman Oct 29 '15 at 19:25
• I think that allowing the input to have deltas of more than 1 would potentially require the hikers to make fractional steps, which would certainly complicate the spec. I don't think the bonus has any point: if you look at the many many existing questions about reachability in a graph (which is what this is), I think you'll find that most of the answers do breadth-first search and thus produce optimal solutions. – Peter Taylor Oct 29 '15 at 22:21
• @PeterTaylor: Yeah, I realized last night that delta > 1 is a problem in the discrete case (if I want hikers to always be at integer steps). You're probably right about the bonus being pretty much useless. – El'endia Starman Oct 29 '15 at 22:24

Bridge Flowers code-golf

One of the example inputs for this challenge is a really minor puzzle in Undertale, and the example output is the solution. You have been warned.

In the Underground, there is a mysterious plant known as the Bridge Flower. They always come in groups of four Bridge Seeds, and have some interesting properties:

• When dropped in water, a Bridge Seed floats away from you in the direction you were facing when you dropped it. It floats until it hits a wall or another Bridge Seed, at which point it stops moving.
• If four Bridge Seeds are touching, forming a straight line, and none of them are moving, they will bloom into Bridge Flowers. Bridge Flowers can support a monster's weight (or a human, but when will that ever happen?), hence the name.
• Another mysterious plant, the Bell Blossom, has a special interaction with Bridge Flowers. The Bell Blossom can be rung to turn all Bridge Flowers back into Bridge Seeds and return them to their original spot.

Consider this map, where . is ground, is water, and X is a Bridge Seed. You'll always be on the part of the map where the Bridge Seeds are. The gap is only four tiles wide, so it's quite easy to make a bridge of Bridge Flowers.

........    ........  |  ........    ........
..X.....    ........  |  ........    ........
.X.X....    ........  |  ........XXXX........
..X.....    ........  |  ........    ........
........    ........  |  ........    ........


Here's a slightly more difficult gap to cross. Even though the straight-line distance is longer than four tiles, you can make it across with a little bit of thought.

........       ........  |  ........       ........
..X.....       ........  |  ........       ........
.X.X....       ........  |  ........       ........
..X.....       ........  |  ........       ........
........       ........  |  ........       ........
..         ............  |  ..     XXXX............
..         ............  |  ..         ............


Here's one that's even trickier than the last two. Most of this river is three tiles wide or fewer, so you can't simply make a straight-line bridge across! Thinking outside the box, though, you'll arrive at the solution.

....   ............  |  .... X ............
...   .............  |  ...  X.............
....   ............  |  .... X ............
.....  ............  |  .....X ............
....   ...X.X......  |  ....   ............
....   ...X.X......  |  ....   ............
..............  |       ..............
............  |         ............


For this challenge, write a program or function that takes a map like the ones here and outputs the same map, but with the puzzle solved (the Bridge Seeds moved to the final position of the Bridge Flowers). If there's more than one possible answer, any of them will do. You may assume there will be exactly two areas of land separated by water, with the one that you start on denoted by the presence of exactly four Bridge Seeds.

If the program can't find an answer... well, just because it's impossible doesn't mean you should give up! Don't lose your Determination! Output Ding to sound the Bell Blossom, reset the Bridge Seeds and try again!

• It took me quite a while to understand the examples. At the very least I think you need to explain that you can pick up bridge seeds from the ground, and clarify "floats until it hits a wall or another Bridge Seed" to define hits. (I think it must include being adjacent to in directions perpendicular to the direction in which it's floating - or am I missing something?) It might also be worth stating explicitly that you don't have to drop them all from the same point. – Peter Taylor Nov 5 '15 at 22:07

Based on this question in the Puzzling SE. Not 100% sure on the scoring, and this is my first time posting a PPCG challenge, so I'd appreciate any suggestions on improving it.

Unicode Box Drawing

Write a program which outputs a diagram using all of the standard Unicode box drawing characters:

─ │ ┌ ┐ └ ┘ ├ ┤ ┬ ┴ ┼
═ ║ ╔ ╗ ╚ ╝ ╠ ╣ ╦ ╩ ╬
╒ ╕ ╘ ╛ ╞ ╡ ╤ ╧ ╪
╓ ╖ ╙ ╜ ╟ ╢ ╥ ╨ ╫


The diagram should have no loose ends, so:

┌┐ <-Allowed  ┌┐ <-Not allowed  ┌┐┌┐ <-Allowed
└┘            └┴                └┴┴┘


Also, single lines should line up with other single lines, and double with double:

├─, ╟─, ╠═, ╞═ <-Allowed  ├═, ╟═, ╠─, ╞─ <-Not allowed


The diagram must use each of the 40 characters at least once, and must form one contiguous, connected shape. Because the characters might look different depending on your browser font, do not consider the space between double lines to be an actual separation. Separations can only be between two characters. In other words:

┌╖ <-Allowed  ┌┐┌┐ <-Not allowed
└╜            └┘└┘


Your code must compute the diagram from scratch. You may not create an already existing diagram and just print it out. A bonus of 20% will be awarded if your program outputs a different valid diagram based on a random seed. (This seed may be provided as input, or you may use system time, read the seed from a file, or any other means you choose.) Note that a random seed may be provided externally, but the diagram itself may not. If you implement this bonus, please provide multiple examples showing different output.

This is code-golf, so scoring will be based on the number of bytes in the source code. In addition, the length of the output will also be counted. You get 40 characters free to account for the required characters. Any additional characters will come at a 10% penalty. This includes spaces/tabs, and any duplicate graph pieces. Newlines are free (to be fair to the differences between OSs).

If your program produces multiple outputs of different lengths (due to randomness), you may use the shortest output produced, but you must provide a seed that results in that output so it can be verified. Since this might be difficult if you used system time as a seed, entries using system time must somehow prove that they are capable of producing the smallest claimed output. (For instance, if they produce the same length output every time, or produce the minimum claimed output at least 50% of the time.)

So in summary:

• Take length of source code in bytes.
• Add 10% for each whitespace character (not counting newlines) or duplicated character in the output.
• Subtract 20% if you implement the random seed input.

Standard loopholes apply.

• "You may not create an already existing diagram and just print it out." This is probably too fuzzy. There many shades of hardcoding. Starting from including the string plainly in the code, over decoding a Base64-string which magically happens to give the correct result, all the way to "cleverly" pruning the possible diagrams your code tries to find, because you know there is a valid one in the search space. You can work around all of that by requiring programs to take some input: e.g. the actual set of characters that need to be accounted for. – Martin Ender Nov 4 '15 at 15:22
• Additionally, this might make an interesting code challenge instead of code golf. – Martin Ender Nov 4 '15 at 15:22
• You should probably also define contiguous, because some of the answers on the Puzzling.SE question have shapes which don't look contiguous (but are considered contiguous) because they have gaps inside the double-line characters (e.g. the "separate" box in the top left corner in the second example of the accepted answer). – Martin Ender Nov 4 '15 at 15:24
• @MartinBüttner This problem is caused by different rendering of the unicode characters. Here is how it looks in the SE Android app: i.stack.imgur.com/PlQWI.jpg – Sleafar Nov 4 '15 at 16:05
• @Sleafar Oh, interesting. Might be worth mentioning anyway if some fonts display it differently. – Martin Ender Nov 4 '15 at 16:07
• Yeah, most fonts I've seen they look like they do in Sleafar's image. For whatever reason, the one used on SE does not. I think it would probably be much harder if you considered those double lines to be actual separations. As for the hardcoding, I didn't want this to be just another "Print out <blah>" challenge, there's tons of those. It's much more interesting if you have to derive such a graph in code. (Which is why I added the bonus for using a random seed, because then you KNOW it's being generated on the fly.) – Darrel Hoffman Nov 4 '15 at 17:56
• To avoid the hardcoding issue, I'd suggest generalizing the puzzle in some way, perhaps by giving a multiset of components to use. – xnor Nov 4 '15 at 21:51
• Problem with using a multiset of input is you'd now have to validate said input to make sure that it's even possible to create a graph with no loose ends using the given inputs. I think the idea of using an input parameter which is just a random seed should be sufficient to ensure the results were not hard-coded in some way. Also, there's nothing stopping you from hard-coding it and printing the same thing every time, you just won't be eligible for the 20% bonus that way. (Is 20% enough? I feel like generating a valid graph is orders of magnitude more difficult than just printing one...) – Darrel Hoffman Nov 4 '15 at 22:52
• You could take the multiset as input and require that at least the characters in it have to be used, possibly others. The scoring could combine source code size and output size for a set of test cases. – Zgarb Nov 5 '15 at 2:16
• I don't like the idea of using preset test-cases except as examples. For one, it makes it possible to hardcode valid graphs for just the test-cases, which is contrary to the spirit of the challenge. Also it limits the challenge to a few already-solved cases instead of making it more general. As Sleafar found in his answer in puzzling, there are many perfect solutions for a given set, and I wouldn't want everyone to just use the same algorithm and thus produce the same graph. I could see providing some examples, but I don't want to use only the example input for scoring purposes. – Darrel Hoffman Nov 6 '15 at 14:21

Find pattern in pixel carpet

[insert story why this needs to be as short as possible]

Given an image the goal is finding the biggest* connected* pattern of pixels, that occurs more than once. If there is more than one distinct pattern of maximum size, find the one that occurs more.

The output should consist of mask (bw image) of the size of the input, that is black where the copies of the pattern were on the original image, and white otherwise.

* Specs

• Biggest as in number of pixels
• connected means four connected

Two instances of the pattern must not overlap.

Meta

• consider rotations / mirroring of the patter too, or only translations?
• runntime requirement?
• "You are a carpet salesman who is looking for stains in a carpet to maintain your reputation. However, since you are locally known as the 'guy who sells dirty carpets', you don't have the money to buy new hardware, and you only have a tiny hard drive and a terrible camera to try to find them with. Therefore, you're going to need a really small program in order to try to find these stains." – Addison Crump Nov 7 '15 at 18:53
• @VoteToClose Then it must be runnable in embedded system? A Raspberry Pi, maybe. – Xwtek Nov 16 '15 at 15:18
• @ChristianIrwan Let's say it is an arduino=) – flawr Nov 16 '15 at 20:32

Optional parentheses

In text, parentheses can denote optional asides. These (often intrusive) additions can be read or skipped over, either option giving a valid sentence. Given a string with parentheses, possibly nested, output all possible readings where each parenthesized section may be omitted.

The(( quick)( brown)) fox jumps( over the( lazy) dog).

The fox jumps.
The quick fox jumps.
The brown fox jumps.
The quick brown fox jumps.
The fox jumps over the dog.
The quick fox jumps over the dog.
The brown fox jumps over the dog.
The quick brown fox jumps over the dog.
The fox jumps over the lazy dog.
The quick fox jumps over the lazy dog.
The brown fox jumps over the lazy dog.
The quick brown fox jumps over the lazy dog.


The outputs can be in any order, but must be distinct. Note the case of (( quick)( brown)), where omitting both quick and brown is redundant with omitting the whole expression. But, the input won't contain any parens that enclose the empty string or are redundant with another pair of pairs.

You may not use regular expressions.

The input string will consist of only letters, spaces, punctuation .,, and parens (). The parens will be properly matched.

• So the example is an invalid input? – feersum Nov 17 '15 at 10:20
• @feersum It's valid, but in trying to clarify the redundancy condition, I realized it's unclear what ( quick)( brown)( quick brown) should give. Hmm. – xnor Nov 17 '15 at 10:23
• "You may not use regular expressions." T_T – Martin Ender Nov 17 '15 at 11:20
• @MartinBüttner Would it make you happier if I made reg-exp and non-reg-exp two separate "leagues"? – xnor Nov 17 '15 at 11:21
• Hm, I don't know. I wonder if it's a bit of a duplicate of this. If I was allowed to use a regex approach, my answer would be almost the same (instead of repeatedly moving the contents of the parentheses to a new line, I'd duplicate the line and remove the parentheses from one of the copies, and the parentheses together with their contents from the other; followed by duplicate removal). – Martin Ender Nov 17 '15 at 11:24
• (Also note that even Retina can't beat a non-regex approach in Pyth on that challenge, so I doubt it'll be much different here.) – Martin Ender Nov 17 '15 at 11:25

Compete with awk

The goal of this challenge is to see if the assertion often see in question on SO is true:

This would be faster in whatever

Test Input

A 10 million lines file generated by this script line:

awk 'BEGIN{for (i=1;i<=10000000;i++) print (i%5?"miss":"hit"),i,"  third\t \tfourth"}' > file


Head of it:

$head -10 file miss 1 third fourth miss 2 third fourth miss 3 third fourth miss 4 third fourth hit 5 third fourth miss 6 third fourth miss 7 third fourth miss 8 third fourth miss 9 third fourth hit 10 third fourth  Goal: The main goal is to compete with awk, to benchmark it will be on this challenge description: • Split each line on any number of spaces (regex [ ]+) • If the line start by hit X with X being an even number • print fields: 4th 1st and 3rd (in this order) Example output (first 5 lines): fourth hit third fourth hit third fourth hit third fourth hit third fourth hit third  Restrictions: The idea being to compete with awk, your program must behave on the same way: • The file must be read. • The match and output order must be modifiable as an awk program can be (Needing a simple code update is OK, it has not to be a command line parameter). • The number of line in input file should be considered unknown. • As long as your output can be piped to awk, the submission is valid. Validation: Pipe the output of your program to awk '!seen[$0]++{unq++;r=$0} END{print ((unq==1) && (seen[r]==1000000) && (r=="fourth hit third")) ? "PASS" : "FAIL"}'  If it show PAST, you're in :) Winner The fastest code would win, to bench yourself here is a mawk version really competitive (which won't count toward answers as I think it's pretty hard to beat): awk '/^hit +[0-9]*[02468] / { print$4, $1,$3 }' file


I've build a test suite on github here which build a result table here.

Answers will be integrated within it to bench all answer on the same machine, if there's a specific way to launch your program, make it appear in the answer so I don't penalize your answer by the way it is launched.

Side challenge:

If you wish to save me time and craft a pull request to include your code in the test system, you're welcome.

• 1. "Split each line on spaces" would IMO split the line miss 1 third fourth into 15 fields. A possible rewording might be "split each line around the regex [ ]+" (where the otherwise unnecessary [] prevent MarkDown from trimming the space). 2. The question as worded is a bit too special-cased, as is the awk code you provide. If it's permissible to optimise the regex to /hit [0-9]*0/ because you know a priori the form of the input, why would it not be permissible to optimise further and write print "fourth hit third\n"*1E6? – Peter Taylor Nov 19 '15 at 21:02
• @PeterTaylor Thanks for the feeback, I reworded a little to be more specific on restrictions and avoid too much test input optimizations. – Tensibai Nov 20 '15 at 10:20
• I still think your reference answer is cheating with /hit [0-9]*0 /. To match the challenge description it should use /^hit +[0-9]*[02468] / – Peter Taylor Nov 20 '15 at 11:18
• @Peter ok, it does no difference in timing at end anyway (at least nothing I've been able to see on slow machines when benchmarking the codes present on GH) – Tensibai Nov 20 '15 at 11:24

Count up by factor keys

TODO: Better title

Information

Let the factor key of a number n be the result of the following process:

• Compute the prime factorization of n: n = p1e1 * p2e2 * p3e3 * ... where all pi are prime, distinct and in increasing order, and all ei are positive integers.

• Compute each piei; call these factors the fi. So n = f1 * f2 * f3 *....

• Convert each fi to a string.

• Concatenate all the strings together.

• Evaluate as a number.

Examples:

The factor key of 2376 is 82711, since 2376 = 23 * 33 * 11 = 8 * 27 * 11.

The factor key of 931 is 4919, since 931 = 72 * 19 = 49 * 19.

As a special case, the factor key of 1 is 0.

Challenge

Output all positive integers, starting from 1, in increasing orders of their factor keys. If two numbers have the same factor key, output the smaller number first. The output must be separated by a delimiter, but this delimiter can be any non-numeric character. If you choose spaces, your output should start with the following:

1 2 3 4 5 7 8 9 11 13 16 17 19 6 23 10 25 14 27 29 31 37 41 12 43 28 47 36


This is , so the shortest solution in bytes wins.

• In the case (I'm not sure whether it occurs, but you haven't stated that it doesn't) that two numbers have the same factor key, can they be output in any order or do you want to define a tie-break? – Peter Taylor Nov 30 '15 at 12:35
• @PeterTaylor It does occur (see 36 and 49), but the question states "If two numbers have the same factor key, output the smaller number first." – user45941 Nov 30 '15 at 20:48
• By reverse engineering, it seems that the e_i must be non-zero. That's worth stating in the spec, and backing up with an example (i.e. any odd number greater than 1). – Peter Taylor Nov 30 '15 at 22:18
• @PeterTaylor Done. – lirtosiast Nov 30 '15 at 22:25

This is currently WIP, but please post feedback if you have any!

Best General-Purpose Compression

Abstract

This paragraph is meant to summarize, the binding rules are below.
Your task is to implement the best general purpose compression algorithm in the context of this challenge. You will be given a file corpus which contains a mix of different file types. The goal is to beat an existing answer in terms of overall size. The size is determined by the size of the compiled compression program (and decompression program if they are two different programs) plus the compressed size of the corpus. Your algorithm must beat standard 7z with default settings by at least 10 bytes, or if there is an existing answer, this answer by 10 bytes.

Corpus

This paragraph lists all files that are part of the input to your program, their size and the baseline 7z score in 7z 9.38 beta. Licenses are linked where needed. Disclaimer: There is no political or humorous motivation behind the choices. All files were chosen because of their representation of certain formats or unique compression behavior.

Natural Language

#1 - ASCII - Bibliography of Computer Security Articles 1983-88: DownloadSave-As

Uncompressed:      146 119
7z 9.38 beta:       17 142
File name:    articles.law


#2 - ASCII - Pre-Flight Launch Preparation for the Space Shuttle: DownloadSave-As

Uncompressed:        44 515
7z 9.38 beta:        15 304
File name:    6_2_4_4_2.TXT


#3 - ASCII - 1984 News report archive about the Bundespost Hack: DownloadSave-As (transcribed by Chaos Computer Club, cites full-text articles from Die ZEIT, taz et al.)

Uncompressed:      102 242
7z 9.38 beta:       32 382
File name:    boh-20f8.txt


#4 - ASCII - (Unofficial historical transcript of) His Last Bow by Arthur Conan Doyle: DownloadSave-As (lic. Public Domain)

Uncompressed:           330 042
7z 9.38 beta:           110 481
File name:    doyle-his-382.txt


#5 - UTF8 - Transcript of the Rosetta Stone translated to English: DownloadRAW->Save-As (lic./© The Nile, Notes for Travelers in Egypt, by E. A. Wallis Budge, 9th Edition, London, Thos. Cook and Son, [1905], pp. 199-211)

Uncompressed:                  18 270
7z 9.38 beta:                   6 299
File name:    Rosetta_EN_UTF8+BOM.txt


Source Code (+ Markdown)

#1 - C - CPU control code taken from the linux kernel: DownloadRAW->Save-As (lic. GPL, Rusty Russell)

Uncompressed: 18 866
7z 9.38 beta:  5 683
File name:     cpu.c


#2 - QBASIC - COMBAT ARENA (Game): DownloadRAW->Save-As (© Jeff Copperthite)

Uncompressed: 59 792
7z 9.38 beta: 11 921
File name: COMBAT.BAS


#3 - XML - The 9/11 Wikimedia dump: Download (lic. GFDL, CC-BY-SA 3.0)

Uncompressed:                                40 557 621
7z 9.38 beta:                                   578 993
File name:    sep11wiki-20071116-pages-meta-history.xml


#4 - C# - Source of the .NET Framework System.Array (CLR): DownloadRAW->Save-As (lic. MIT)

Uncompressed:  133 646
7z 9.38 beta:   15 172
File name:    Array.cs


#5 - JavaScript - jQuery 1.11.3: DownloadRAW->Save-As (lic. MIT)

Uncompressed:          284 394
7z 9.38 beta:           73 585
File name:    jquery-1.11.3.js


Binary Media

#1 - JPG - "Flower in the Garden": Download (lic. CC-BY-NC-ND 2.0 & © wellenkern)

Uncompressed: 10 482 521
7z 9.38 beta: 10 434 631
File name:    flower.jpg


#2 - MP3 - "Mariam's Cake" by Robin Grey: Download (lic. CC-BY-NC-SA 3.0)

Uncompressed: 8 468 990
7z 9.38 beta: 8 162 146
File name:     cake.mp3


#3 - Bitmap24 - DP1M7218 16:9 COLOR: Download (lic. CC-BY-SA 3.0 & © iKobe!)

Uncompressed: 37 834 326
7z 9.38 beta: 14 652 746
File name:    street.bmp


#4 - MP4 - "Home 5x5": Download (lic. CC-BY 3.0 & © Michael Jones)

Uncompressed: 6 173 898
7z 9.38 beta: 6 013 047
File name:     home.mp4


#5 - binary - Block of cryptographically secure random data, a.k.a. "The Incompressible File": Download

Uncompressed: 100 000
7z 9.38 beta: 100 129
File name:    rnd.bin


Rules

(De/)Compression Program and Algorithm

1. The program used to compress and decompress a file is the compiled executable (or executable script).
2. You may use two or more programs and ship additional files if needed, however only one program per step (a step is de- or compression) is executed.
3. No program or part of a program is allowed to request resources that are not included in your distribution. Exceptions are standard system resources (clean install, i.e. a Java runtime would not be available when you use Windows).
4. You may use linux-, windows- or architecture-specific code.
5. You must not modify files or file names. The file names listed above are binding. Note: Some downloads may not carry the file name, you have to rename them (case-sensitive). The file names of compressed files do not matter. However, the original file names must be restored after decompression.
6. You can use any existing or new algorithm. Content mixing is of course allowed, however, you may not call any compression routines available on your system (or language). You program must be a self-contained implementation.
7. Your algorithm must not specifically target this corpus, it has to be general purpose. You are however allowed to target the used file types (e.g. by identifying the type via "Magic Bytes" and adjusting the algorithm accordingly).
8. The (de)compression program only accepts one argument, a file name. Piping program output somewhere or providing anything else is forbidden. (I.e. you'll have to write the output file yourself).

Running your entry

Your compression program is called 15 times, once for every file in the corpus:

comp articles.law
comp ...


This process must produce 15 compressed files on disk in the same directory (original files can be overwritten). (comp is just an example name, you don't have to use this name)

Decompression is similar. Your decompression program (which may be the same executable) is called 15 times, once for every file in the corpus:

decomp ...


Where ... is the file name of the compressed file. This will produce the original 15 files on disk (with their original file names) (compare the SHA256 when in doubt).

Neither the compression or decompression program may take any additional command line parameters.

Scoring

Your score is P+C, where

• P is the size of all programs used in your distribution
• C is 15 files, the compressed corpus

in bytes.

Your answer must include the following:

• The source code for all parts of P. (Is not counted in the score).
• A link to a GitHub repository containing both P and C. Means you actually have to run your solution and include all compressed files. If your repository features additional files that should not be included in your score (gitignore, README etc.), put the scored files in a separate folder.

To beat someones score, your score has to beat it by at least 10 bytes. If some answer has the score X, then your answer has to have a score of N <= X-10, where N is your score. The starting score is set by 7z.

Invalid Answers

Please refrain from posting invalid answers, such as:

• Answers without source code for P,
• answers without compressed files, or where compressed files run against P produce anything but the original corpus,
• answers that contain only partial solutions.

For clarity, here's how an answer might look like:

I did this algorithm X. This code is used for compression:

<comp.c>

and this for decompression:

<decomp.c>

I used windows, my score therefore is 9001:

comp.exe:          234 bytes
decomp.exe:        124 bytes
Compressed files: 8643 bytes

You can download the programs and compressed files from my
repository here: github.com/foobar42.


Leaderbaord and History

Current score: 423,424 (P) + 40,229,661 (C) = 40,662,085
Set by: 7z 9.38 beta (Win32) on 2015-12-18

• I love this, even if it's above my skill level. I hope it gets posted! – cat Dec 18 '15 at 3:18
• 1. "The file names listed above are binding" for what? My best guess is that the compressor has to store the original filename in the file, and the decompressor has to output to that name, but I didn't notice that that was stated explicitly anywhere. 2. comp articles.law: are we required to use command-line arguments and to pick an output filename ourselves, or could we use stdio and invoke as comp <articles.law >articles.law.myz? 3. "Your answer must include ... A standard ZIP file". How? – Peter Taylor Dec 18 '15 at 8:38
• @PeterTaylor 1. It is irrelevant how the file name is stored, however it must be restored after decompression. I'll add that. 2. The compressed filename is not specified in the rules so it can be anything. According to the current rules comp <articles.law >articles.law.myz would be invalid. The only calid way to call one step is program file, files must be created by the step program. I'll add that. 3. I'm thinking of changing this part to "GitHub Repository". Suggestions? – mınxomaτ Dec 18 '15 at 16:12
• I would posit that there is no such thing as truly "General-Purpose Compression", at least in an information-theory sense. Text (and various other formats) compress nicely because they are not general. But compression of random data will not give any gain in the long run. – Digital Trauma Dec 18 '15 at 20:47
• @DigitalTrauma GP in terms of compression algorithm means that the algorithm does not specifically target the specific files in a test suite. ZIP for example is a GP algorithm, as most other algorithms. – mınxomaτ Dec 18 '15 at 20:53
• Rather than using GitHub repos, I think the best approach is to require all answers to be deterministic and then the scores can be verified by just running them. – Peter Taylor Dec 19 '15 at 12:26

Add two numbers (or Arithmetic on continued fraction expansions)

Given two continued fraction expansions N,M as lists of integers (in any suitable (human readable) format for your language) and an operation @ (addition, substraction, multiplication and division) +-* or / (e.g. encoded as numbers 1-4 or as characters or whathever is suitable for your language) return N@M (the sum/difference/product/ratio) again in the form of a continued fraction expansion.

By continued fraction expansion I mean a simple continued fraction expansion which are of the form:

                   1
a0 + -----------------------
1
a1 + ------------------
1
a2 + ------------
1
a3 + ------
...



Which will be represented as a list of the form [a0,a1,a2,a3,...]

Meta

Is this too complicated? Should the challenge be reduced to only one operation e.g. + or *?

• It's not very complicated, and handling the four functions isn't much harder than handling just one; but if you assume that the input is two finite lists then it's possible to do a boring approach of evaluating them as reduced rationals, doing the operation, and converting back. IMO it would be more interesting to restrict to languages which support either functions with state or infinite lazy lists and to include some test cases which work with "nice" irrationals like sqrt 2 and e evaluated to a given number of terms. – Peter Taylor Jan 2 '16 at 17:33
• Would that be equivalent to letting the program take two lists of sufficient* length and a number n, and requiring it to return the first n terms? *sufficiently long such that the first n terms can correctly be evaluated. – flawr Jan 2 '16 at 19:01
• I don't see how that would avoid the approach of converting to rationals, performing the operation, and converting back. – Peter Taylor Jan 2 '16 at 19:50
• I am just trying to understand what you meant in the first comment=) You obviously can only output a finite number of terms, but this also implies that you only need a finitely many input terms. So I do not see why you would have to use 'infitine' lists as input, if you can only output finite lists. And for not excluding too many languages, my suggestion was passing another argument n that determines how many terms the output should consist of. – flawr Jan 2 '16 at 21:57
• It's the difference between "You only need these a terms" and "You need a terms, but it's up to you to work out what a is". Working out how many terms you need is the interesting part; converting between a finite continued fraction and a rational expressed as numerator and denominator is comparatively boring. – Peter Taylor Jan 2 '16 at 23:21
• Now I see what you mean, thank you for elaborating! I like your suggestion, I'm going to try to come up with some suitable test cases. – flawr Jan 3 '16 at 13:50

Number of cells used in optimal BF representation of a constant

Your task is deceptively simple. Given a number n (or a character C with char code value n in your encoding), output the number of cells utilized in the optimal representation of that number n in a standard Brainf*** interpreter with 0 <= n <= 255. (8-bit wrapping.)

Examples

> input
< output

> 4
< 1   ; ++++

> 0
< 1   ; empty program

> 201
< 4   ; >+[-->-[-<]>]>+

> 190
< 5   ; ++[-->-[<]>-]>

> 185
< 2   ; -[>-<-------]>++

> 255
< 1   ; -

> 242
< 1   ; --------------

> 241
< 2   ; ---[>-----<+]>


Meta

Does more describing need to be done? Should I just remove the bonuses? Bonuses removed.

• "Optimal" in what sense? And surely 0 needs one cell, to store the output? – Peter Taylor Jan 7 '16 at 19:25
• I too think that this needs clarification. In the chat you said optimal representation = the shortest brainfuck program outputting this number. But I think there still might be some ambiguity, as the shortest might not be unique, and therefore might not have an unique number of used cells. – flawr Jan 7 '16 at 21:03
• @flaws that is true. Minimal output length, minimizing the number of cells used. – Conor O'Brien Jan 7 '16 at 21:04
• @PeterTaylor optimal as in shortest. Will add that at home. And I'm referring to the number of cells required to modify another cell. Perhaps I'm wrong in my reasoning. – Conor O'Brien Jan 7 '16 at 21:06
• However you count the cells, 0 should surely return the same value as 4, because neither of them use <>. – Peter Taylor Jan 7 '16 at 21:41
• I vote for "remove bonuses" – quintopia Jan 9 '16 at 6:37
• @quintopia Agreed, in retrospect they are a bad idea. – Conor O'Brien Jan 9 '16 at 17:38
• @ThomasKwa Indeed. – Conor O'Brien Jan 9 '16 at 17:38
• @PeterTaylor That is true. – Conor O'Brien Jan 9 '16 at 17:39
• I'm not sure what part of this challenge is "deceptively simple". ;) – Martin Ender Jan 10 '16 at 15:01

Roll a Die

Most of the time in PPCG, challenges that involve dice rolling normally focus on some numeric property. However all I want you to do is to actually display the die itself in an isometric 3D format.

Your program or function should accept two optional integers representing the numbers to be displayed on the top and the front of the die respectively. If no valid value for the top is given then one of the 24 possible rolls should be uniformly chosen (but it does not need to be completely random so a time-dependent result is acceptable). If no valid value for the front is given then one of the 4 valid values should be uniformly chosen.

The die to be displayed is a standard Western die. Example: when rolling a 4, if the 1 is at the front, the number to the right is a 5; to display any other number would be an error. A more complex example is also shown below; in this case, not only is the 2 the correct number to display on the right, but all three numbers must be displayed as shown, i.e. the diagonal of the number 2 must point from 1/4 to 3/6 and not 1/3 to 4/6; the diagonal of the number 3 must point from 1/5 to 2/6 and not 1/2 to 5/6; the number 6 must point from 3 to 4 and not from 2 to 5.

    o-------o     o-------o
/ *   * /|    / *   * /|
/       /*|   / *   * / |
/ *   * /  |  / *   * /  |
o-------o*  | o-------o*  |
|       | * | |     * |   |
|       |  *o |       |  *o
|   *   |  /  |   *   |  /
|       |*/   |       | /
|       |/    | *     |/
o-------o     o-------o


You must draw all the sides, corners and pips preferably using the characters as shown above. Additional whitespace is permissible if you are consistent (only the pips may vary between rolls).

This is , so the shortest answer will be the winner.

Reference code (Batch):

@echo off
setlocal enableextensions
call :%1%2 2>nul
if errorlevel 1 call :%1 2>nul
if errorlevel 1 goto %time:~6,1%
goto :eof
:1
set /a front = (%time:~7,1% %% 2) * 2 + (%time:~9,1% %% 2) + 2
goto 1%front%
:2
set /a front = (%time:~7,1% %% 2) * 3 + (%time:~9,1% %% 2) * 2 + 1
goto 2%front%
:3
set /a front = (%time:~7,1% %% 2) * 4 + (%time:~9,1% %% 2) + 1
goto 5%front%
:4
set /a front = (%time:~7,1% %% 2) * 4 + (%time:~9,1% %% 2) + 1
goto 5%front%
:5
set /a front = (%time:~7,1% %% 2) * 3 + (%time:~9,1% %% 2) * 2 + 1
goto 5%front%
:6
set /a front = (%time:~7,1% %% 2) * 2 + (%time:~9,1% %% 2) + 2
goto 6%front%
:12
echo     o-------o
echo    /       /^|
echo   /   *   /*^|
echo  /       /  ^|
echo o-------o   ^|
echo ^| *     ^| * ^|
echo ^|       ^|   o
echo ^|       ^|  /
echo ^|       ^|*/
echo ^|     * ^|/
echo o-------o
goto :eof
:13
echo     o-------o
echo    /       /^|
echo   /   *   /*^|
echo  /       /  ^|
echo o-------o*  ^|
echo ^|     * ^| * ^|
echo ^|       ^|  *o
echo ^|   *   ^|  /
echo ^|       ^|*/
echo ^| *     ^|/
echo o-------o
goto :eof
:14
echo     o-------o
echo    /       /^|
echo   /   *   / ^|
echo  /       /  ^|
echo o-------o*  ^|
echo ^| *   * ^|   ^|
echo ^|       ^|  *o
echo ^|       ^|  /
echo ^|       ^| /
echo ^| *   * ^|/
echo o-------o
goto :eof
:15
echo     o-------o
echo    /       /^|
echo   /   *   /*^|
echo  /       /  ^|
echo o-------o*  ^|
echo ^| *   * ^|   ^|
echo ^|       ^|  *o
echo ^|   *   ^|  /
echo ^|       ^|*/
echo ^| *   * ^|/
echo o-------o
goto :eof
:21
echo     o-------o
echo    / *     /^|
echo   /       /*^|
echo  /     * /  ^|
echo o-------o*  ^|
echo ^|       ^|   ^|
echo ^|       ^|  *o
echo ^|   *   ^|  /
echo ^|       ^|*/
echo ^|       ^|/
echo o-------o
goto :eof
:23
echo     o-------o
echo    /     * /^|
echo   /       / ^|
echo  / *     /  ^|
echo o-------o   ^|
echo ^| *     ^| * ^|
echo ^|       ^|   o
echo ^|   *   ^|  /
echo ^|       ^| /
echo ^|     * ^|/
echo o-------o
goto :eof
:24
echo     o-------o
echo    /     * /^|
echo   /       /*^|
echo  / *     /* ^|
echo o-------o*  ^|
echo ^| *   * ^|   ^|
echo ^|       ^|  *o
echo ^|       ^| */
echo ^|       ^|*/
echo ^| *   * ^|/
echo o-------o
goto :eof
:26
echo     o-------o
echo    / *     /^|
echo   /       / ^|
echo  /     * /  ^|
echo o-------o*  ^|
echo ^| * * * ^| * ^|
echo ^|       ^|  *o
echo ^|       ^|  /
echo ^|       ^| /
echo ^| * * * ^|/
echo o-------o
goto :eof
:31
echo     o-------o
echo    /     * /^|
echo   /   *   /*^|
echo  / *     /  ^|
echo o-------o   ^|
echo ^|       ^|   ^|
echo ^|       ^|   o
echo ^|   *   ^|  /
echo ^|       ^|*/
echo ^|       ^|/
echo o-------o
goto :eof
:32
echo     o-------o
echo    / *     /^|
echo   /   *   /*^|
echo  /     * /  ^|
echo o-------o* *^|
echo ^|     * ^|   ^|
echo ^|       ^|* *o
echo ^|       ^|  /
echo ^|       ^|*/
echo ^| *     ^|/
echo o-------o
goto :eof
:35
echo     o-------o
echo    / *     /^|
echo   /   *   / ^|
echo  /     * /  ^|
echo o-------o   ^|
echo ^| *   * ^| * ^|
echo ^|       ^|   o
echo ^|   *   ^|  /
echo ^|       ^| /
echo ^| *   * ^|/
echo o-------o
goto :eof
:36
echo     o-------o
echo    /     * /^|
echo   /   *   /*^|
echo  / *     /  ^|
echo o-------o*  ^|
echo ^| *   * ^| * ^|
echo ^|       ^|  *o
echo ^| *   * ^|  /
echo ^|       ^|*/
echo ^| *   * ^|/
echo o-------o
goto :eof
:41
echo     o-------o
echo    / *   * /^|
echo   /       /*^|
echo  / *   * /  ^|
echo o-------o*  ^|
echo ^|       ^| * ^|
echo ^|       ^|  *o
echo ^|   *   ^|  /
echo ^|       ^|*/
echo ^|       ^|/
echo o-------o
goto :eof
:42
echo     o-------o
echo    / *   * /^|
echo   /       / ^|
echo  / *   * /  ^|
echo o-------o   ^|
echo ^|     * ^| * ^|
echo ^|       ^|   o
echo ^|       ^|  /
echo ^|       ^| /
echo ^| *     ^|/
echo o-------o
goto :eof
:45
echo     o-------o
echo    / *   * /^|
echo   /       /*^|
echo  / *   * /  ^|
echo o-------o* *^|
echo ^| *   * ^|   ^|
echo ^|       ^|* *o
echo ^|   *   ^|  /
echo ^|       ^|*/
echo ^| *   * ^|/
echo o-------o
goto :eof
:46
echo     o-------o
echo    / *   * /^|
echo   /       /*^|
echo  / *   * /  ^|
echo o-------o   ^|
echo ^| *   * ^|   ^|
echo ^|       ^|   o
echo ^| *   * ^|  /
echo ^|       ^|*/
echo ^| *   * ^|/
echo o-------o
goto :eof
:51
echo     o-------o
echo    / *   * /^|
echo   /   *   / ^|
echo  / *   * /  ^|
echo o-------o*  ^|
echo ^|       ^| * ^|
echo ^|       ^|  *o
echo ^|   *   ^|  /
echo ^|       ^| /
echo ^|       ^|/
echo o-------o
goto :eof
:53
echo     o-------o
echo    / *   * /^|
echo   /   *   /*^|
echo  / *   * /* ^|
echo o-------o*  ^|
echo ^| *     ^|   ^|
echo ^|       ^|  *o
echo ^|   *   ^| */
echo ^|       ^|*/
echo ^|     * ^|/
echo o-------o
goto :eof
:54
echo     o-------o
echo    / *   * /^|
echo   /   *   / ^|
echo  / *   * /  ^|
echo o-------o   ^|
echo ^| *   * ^| * ^|
echo ^|       ^|   o
echo ^|       ^|  /
echo ^|       ^| /
echo ^| *   * ^|/
echo o-------o
goto :eof
:56
echo     o-------o
echo    / *   * /^|
echo   /   *   /*^|
echo  / *   * /  ^|
echo o-------o*  ^|
echo ^| * * * ^|   ^|
echo ^|       ^|  *o
echo ^|       ^|  /
echo ^|       ^|*/
echo ^| * * * ^|/
echo o-------o
goto :eof
:62
echo     o-------o
echo    / * * * /^|
echo   /       /*^|
echo  / * * * /  ^|
echo o-------o*  ^|
echo ^| *     ^|   ^|
echo ^|       ^|  *o
echo ^|       ^|  /
echo ^|       ^|*/
echo ^|     * ^|/
echo o-------o
goto :eof
:63
echo     o-------o
echo    / *   * /^|
echo   / *   * / ^|
echo  / *   * /  ^|
echo o-------o*  ^|
echo ^|     * ^|   ^|
echo ^|       ^|  *o
echo ^|   *   ^|  /
echo ^|       ^| /
echo ^| *     ^|/
echo o-------o
goto :eof
:64
echo     o-------o
echo    / *   * /^|
echo   / *   * /*^|
echo  / *   * /  ^|
echo o-------o*  ^|
echo ^| *   * ^| * ^|
echo ^|       ^|  *o
echo ^|       ^|  /
echo ^|       ^|*/
echo ^| *   * ^|/
echo o-------o
goto :eof
:65
echo     o-------o
echo    / * * * /^|
echo   /       /*^|
echo  / * * * /  ^|
echo o-------o   ^|
echo ^| *   * ^| * ^|
echo ^|       ^|   o
echo ^|   *   ^|  /
echo ^|       ^|*/
echo ^| *   * ^|/
echo o-------o


        o-------o
| *   * |
|       |
|   *   |
|       |
| *   * |
o-------o-------o-------o-------o
| *   * |       | *     | * * * |
|       |       |       |       |
|       |   *   |   *   |       |
|       |       |       |       |
| *   * |       |     * | * * * |
o-------o-------o-------o-------o
| *     |
|       |
|       |
|       |
|     * |
o-------o

o-------o
/ * * * /|
/ * * * /*|
/ * * * /* |
o-------o* *|
| * * * | * |
|       |* *o
| * * * | */
|       |*/
| * * * |/
o-------o

• Re "real standard dice". Both left-handed and right-handed dice exist (depending on culture), so it might be worth mentioning explicitly which of the two it is, instead of giving a single example. "You can earn a bonus for rolling several dice but in this case you must display the dice horizontally." You forgot to mention how much that bonus is worth but I would recommending leaving it out completely. – Martin Ender Jan 25 '16 at 14:47
• It's also sort of implied that two numbers is also valid input (provided they can be on adjacent sides), but you never state how that input is to be handled. Likewise, what if the input has more than two numbers? I'd also prefer a fixed output format instead of letting people choose their own ASCII art. It's bound to derail into arguments of how much visual complexity is sufficient to display the die properly. – Martin Ender Jan 25 '16 at 14:49
• @MartinBüttner Sorry I didn't know about Chinese dice, and the bad wording in some of my paragraphs; I hope this version is better. – Neil Jan 25 '16 at 17:37
• The diagram of two dice isn't quite enough for me to visualise the all the cases, and the batch script isn't the easiest to read either because of the escaping. I think it would be an improvement to give an unfolded net, possibly a diagram showing all the possible pip positions, and if you want to have a full test suite then to rewrite it in JavaScript and make it a Stack Snippet. Since I've already gone to the effort of producing the first two diagrams to evaluate how helpful I find them, I'll edit them in as a footnote and leave you to decide whether and where you want to use them. – Peter Taylor Jan 25 '16 at 21:02

Polyphonic Pitch Detection

Sandbox Notes

• Still in progress. I haven't made any of the test cases or snippets yet.
• I'm not sure what to tag this as. It doesn't quite fit the definition of any existing tags. maybe?
• I'm thinking of making a monophonic version of this challenge as a precursor to this one, since monophonic pitch detection is much easier and this one may be too difficult for most to attempt. Also note that monophonic would have more of an emphasis on perfecting an algorithm, rather than getting it to work at all.

Take an array of samples, and output the frequencies of the waveforms found in the samples.

Despite how simple this may sound, it is actually quite difficult. Even though it has been researched for almost a century, a robust and versatile algorithm for polyphonic pitch detection is yet to be found. Let's look at a simple waveform as an example:

(The X-axis is in 1/440ths of a second, use the "Play Samples" snippet below to hear what it sounds like)

This is a 440hz sine wave. In musical terms, it is middle A or A4. The next image is a 554hz sine wave (or C#5) on top of the 440hz wave:

It looks exactly the same, except slightly "squashed" (and sounds a bit "higher"). It is a major 3rd above A4, which means they sound pleasant when played together, however when you look at the waveform that adding them together creates:

The resulting waveform appears vastly different. To further complicate matters, it changes shape depending on the time. There are other factors like overtones, background noise and the fact that real-world waveforms are more complex than sine waves which also make it tricky. (But the human brain still manages to do this effortlessly!)

Your Task

Receive a list of samples as signed 16-bit integers at a fixed sample-rate of 44100 samples-per-second. The input waveform will contain between 1 and 5 (inclusive) simultaneous frequencies in the range of 100hz to 2000hz.

You must output a list of frequencies detected (in hertz) with up to 2 decimal places of precision.

Test Cases

Each test case is on it's own line. Each line begins with the name of the test case, followed by a semi-colon (;), then the frequencies present in the test case (accurate to 2 decimal places) separated by commas (,), followed by another semi-colon, then the samples separated by commas:

Test Case Name;123.45,67.8,90.12;3,75,1234,56789,4321,-23,-408,-9266,41,0,etc...

Each test case will be exactly 44100 samples (1 second) long.

(link to test case file, will include synthesized waveforms, real instrument sounds, multiple instrument/waveform types, a variety of harmonies and pitches, combinations of each of these)

Scoring

The score of a submission is a percentage based on how close the submission's results are to the actual frequencies of each test case. Specifically it will calculated using the formula in the snippet below (use this to calculate your score):

(snippet for calculating score)

Rules

• No built-ins that detect pitch or extract waveform frequencies are allowed.
• Helper functions that are designed to aid frequency analysis like FFT are permitted.
• You may optimise your solution for the test cases, but you cannot hard-code the results for these specific test cases.

Play Samples

You can hear what certain frequencies or a list of samples sounds like by pasting them into this snippet (requires a browser that supports the Web Audio API):

document.write('Enter frequencies separated by commas: <input type="text" ' +'id="Frequencies" value="440,554"><br>' +'<button id="Play" onclick="togglePlay()">Play</button><br>' +'Paste the whole test case line here: <input type="text" id="Test"><br>' +'<input type="checkbox" id="Loop"> Loop?<br>' +'<button id="PlayTest" onclick="togglePlayTest()">Play</button>');var position = 0,samples = null,sampleSize = 0xffff + 1,halfSampleSize = sampleSize / 2,bufferSize = 4096,sampleRate = 44100,context = new AudioContext(),processor = context.createScriptProcessor(bufferSize, 1, 1),oscillators = null;processor.connect(context.destination);processor.onaudioprocess = function(e) {var buffer = e.outputBuffer.getChannelData(0);if(samples) {var sampleLength = samples.length,loop = Loop.checked;for(var i = 0; i < bufferSize; i++) {position += sampleRate / context.sampleRate;if(loop) position %= sampleLength;buffer[i] = ~~samples[position | 0] / halfSampleSize;}if(position >= sampleLength) togglePlayTest();}else buffer.fill(0);};function togglePlayTest() {if(samples) {samples = null;PlayTest.textContent = "Play";}else {var parts = Test.value.split(";");if(parts.length > 1) {PlayTest.textContent = "Stop - " + parts[0] + " (" +parts[1].split(",").map(function(f) { return f + "hz"; }).join(", ") +")";samples = parts[2].split(",").map(function(n) { return +n; });position = 0;}}}function togglePlay() {if(oscillators) {oscillators.forEach(function(o) { o.stop(); });oscillators = null;Play.textContent = "Play";}else {oscillators = [];var frequencies = Frequencies.value.split(","),gain = context.createGain();gain.gain.value = 1 / (frequencies.length + 1);gain.connect(context.destination);frequencies.forEach(function(frequency) {var oscillator = context.createOscillator();oscillator.frequency.value = frequency;oscillator.connect(gain);oscillator.start();oscillators.push(oscillator);});Play.textContent = "Stop";}}

Links

• I think it is patented in some countries, but a quite successfull one is the yin algorithm. And you might want to add test-battery – flawr Jan 18 '16 at 13:27
• @flawr Added. Although note that while the YIN algorithm is quite successful at finding the fundamental frequency, most of the test cases will contain multiple frequencies. – user81655 Jan 19 '16 at 0:01
• Is that desmos I see? – Cyoce Jan 30 '16 at 7:46
• @Cyoce Yep. ;-) – user81655 Jan 30 '16 at 10:38

Anything but stacks

Stack-based languages are, like totally, all the rage these days. From GolfScript, to CJam, to MATL, it seems like stack-based languages are popping up everywhere. But now, it is time to push forward and explore new memory models.

In this challenge, you are to create a new programming language based off of any other data structure.

Still undecided is the "goal" of this challenge. The main idea is that, by using a unique data structure, these new languages may have advantages for certain types of problems. Since different languages would use different structures, however, there's not any single set of challenges that would provide a fair comparison. An alternative idea is to pick a single data structure and have everybody use it.

I'm going to work off the "use a specific data structure" idea, since that's really the only thing that can narrow the focus of this question. So, maybe some options are...

• Queues
• Trees (except I'm not sure how a "tree without branching" would be different from a "stack")
• Priority Queues
• Sets, etc.
• Associative arrays (dicts/hashmaps)

The next step might be to find some types of problems to target. Are there any algorithms that are known to be exceptionally difficult to implement with a stack?

• I think a lot of people will follow Pyth (fixed-arity prefix with iteration as a backup) or Jelly (tacit) if the goal is a golfing language. – lirtosiast Feb 1 '16 at 4:30
• This is currently too broad by miles. – Peter Taylor Feb 1 '16 at 8:17
• I definitely agree with Peter. "An alternative idea is to pick a single data structure and have everybody use it." That seems a like a much better idea, and then you need to make sure that that's the only type of memory that's allowed (maybe except for a finite amount, like a handful of registers). Even then, it definitely needs some sort of goal. Requiring TC-ness might also be a good idea depending on what kind of answers you're looking for. – Martin Ender Feb 1 '16 at 10:11
• I think set-based languages would be the most interesting, since sets can't have an order, and most other data structures can implement stacks easily. There are also so many different directions to take with them. – Fricative Melon Feb 1 '16 at 16:47
• Somewhere far down on my list of esolang ideas are two languages using stacked cups as a data structure (I'd like to implement a very minimalistic one that doesn't go far beyond the instructions in that challenge, and a feature-rich one, where you're actually programming two hands that can move cups). Anyway, I'd be willing to offer that idea as a data structure for your challenge. – Martin Ender Feb 1 '16 at 19:14
• @user1657355 HPR is set-based, although the set can contain lists. – Zgarb Feb 1 '16 at 19:22
• Regular expressions. Those are probably hard with a stack based language. Parsing expressions would be hard, I would think, as well. – Conor O'Brien Feb 1 '16 at 20:06
• I was thinking of two deque based languages. – Xwtek Feb 18 '16 at 12:06

Write Hamlet in 1024 bytes

You will write a program, less than 1024 bytes in length, that outputs this text to a file, STDOUT or nearest available equivalent. Since this is clearly impossible, you don't have to output the exact text, just get as close as possible.

To measure how close your output is to the original, use the Python 3 script below. It works by concatenating the original text onto the end of your file, then compressing the result using the lzma algorithm, and then subtracting the compressed size of your file alone. This works because if your file contains a lot of common features with the original then the compression algorithm can take advantage of this to make a smaller file. A more sophisticated version of this idea is called normalised compression distance.

All answers should contain the code, at least a brief explanation of how it works (full explanations are encouraged), and the first 2000 characters or so of its output.

This is . Scores are calculated using the script below. The lowest score wins.

Rules and clarifications

• Your program must be completely self-contained within a single file, taking no input, loading no files, and executing no other programs. Importing libraries is permitted.

• Your program must run deterministically, producing the same output every time

• Your output must contain only printable ASCII characters, tabs and newlines. (That is, characters with codes 32 to 126 inclusive, plus 9 and 10.) The comparison script checks for this.

• Your output must be the same size as the original file, 182581 bytes.

• You may not use any built-ins or library code that provides compression or decompression algorithms. (e.g. lzma, bz2 etc.) It's OK to use them if you can implement them yourself inside the character limit. Base conversion is OK, and libraries implementing data structures such as Huffman trees are OK.

• If for some reason your language or one of its libraries contains a function that outputs some or all of the text of Hamlet, you may not use that function.

• Your program must be written in a programming language, as defined in this answer. This definition must apply to your source code, not just to any compiler/interpreter flags used to run it.

• In the case of any inconsistency in the script's behaviour between machines, the version on my machine is the definitive one. (Python 3.4.3 on a Mac.)

Here is the comparison code to use. It requires Python 3 because of the lzma dependency. (lzma is much better than bz2 or gzip for this purpose.) It requires the text file linked above to be in the same directory with the name ORIGINAL.txt. Run it with a command like

python3 compare.py my_output.txt


[to do: produce and supply a zip file containing this script together with ORIGINAL.txt]

import lzma
import sys

with open("ORIGINAL.txt", 'r') as file:
orig = file.read()

with open(sys.argv[1], 'r') as file:
text = file.read()

def csize(txt):
return len(lzma.compress(txt.encode('utf-8')))

character_codes = {ord(c) for c in text}
valid_codes = set([9,10] + list(range(32,127)))
if character_codes - valid_codes:
print("NOT VALID: file contains a non-printable character")
else:
print( csize(text + orig) - csize(text) )


Technical notes

The compressed size of the original file is 64976 bytes. In an ideal world the original file would have a score of 0, but it actually scores 100. Shakespeare's Macbeth scores 61776, so that should probably be considered a pretty good score.

• One problem I see is dictionary compression should be really powerful here, but prohibitive for any languages that don't have a built-in dictionary (because including one would clearly exceed the byte limit). What about languages that do have built-in dictionaries though? (Jelly's string literals can be dictionary-compressed out of the box, and I think Mathematica can load a dictionary with a single function call, too.) – Martin Ender Feb 1 '16 at 13:35
• @MartinBüttner I guess I personally am OK with languages with built-in dictionaries - it just seems like using the right tool for the job. But if that would be seen as unacceptably biasing the challenge toward a certain set of languages, I'm OK with putting something in to prohibit it. – Nathaniel Feb 1 '16 at 13:39
• I just realised that the optimal solution might consist of a 1024-character string literal. For that reason I might add a requirement that the output be the same length as the original file. (But first I have to think about whether that's really likely to be a problem.) – Nathaniel Feb 1 '16 at 13:41
• Does base conversion count as a "compression or decompression algorithm"? What about Huffman tree-related functions? – Peter Taylor Feb 1 '16 at 22:32
• @PeterTaylor I think I would say both are OK - I'm mostly just trying to prevent solutions that don't do much other than bz2.decompress('[some long string]'). If you implemented your own decompression then we learn something from the answer, even if you did it using tools designed for that purpose. However, if you have an intuition that this would devolve into something uninteresting please let me know. – Nathaniel Feb 2 '16 at 4:57
• @PeterTaylor I've added a restriction on the source character set as well, which should head off some of the most annoying consequences of those things. – Nathaniel Feb 2 '16 at 5:01
• I asked about base conversion because it's one of the techniques most frequently used on this site for compression, and about Huffman because it seemed like a gray area which should be mentioned explicitly. On an unrelated note, would it not be a better measure to do the concatenation the other way round, so that the "compressed" Hamlet is training the LZMA bigraph dictionary for the real one? Otherwise the best score would probably be obtained by repeating some large chunk from near the end of the real one as many times as needed to meet the target length. – Peter Taylor Feb 2 '16 at 12:50
• @PeterTaylor I've made that change. (I tested it on a few test files, including one composed of the last 5% of Hamlet repeated 20 times, and it didn't make much difference, but it changed the scores by 100 or so.) – Nathaniel Feb 2 '16 at 12:57
• (One of the reasons that LZMA is good for this purpose is that it's close to being a "normal compression algorithm", which includes being close to invariant with respect to swapping the concatenated strings around.) – Nathaniel Feb 2 '16 at 13:07
• I changed my mind about trying to keep the source printable. It's a nice thing to have but it means having rules that are too complex, and people will just try to loophole their way around them anyway. So that means base coding is just fine. – Nathaniel Feb 2 '16 at 13:23

Create an unkillable Windows process

In various versions of the Microsoft Windows operating system, it is possible for a process to enter a state where it cannot be killed by the 'End Process' feature in the Task Manager. Your goal is to create a user-mode program that enters such a state (or spawns a process that does) using as few bytes of code as possible. Please state the version of Windows on which you have tested the program.

Some techniques for creating unkillable processes can be found here.

Be warned that such processes may cause a performance impact, so be prepared to reboot if necessary when testing any of these programs.

TODO:

• Minimum version? The #1 in the accepted answer seems too easy.
• #2 is also lame so maybe taskkill should be the standard instead of Task Manager.
• This is coming fairly close to a challenge asking for malicious software. – Martin Ender Feb 22 '16 at 9:49
• @MartinBüttner It's not completely clear, but the discussion on that question seemed to indicate that a process in a zombie state could not actually execute any longer, making it useless for malware. – feersum Feb 22 '16 at 9:53

The greatest power...

Yeah, I'm still trying to come up with a better title...

A positive integer n is a perfect power of order k if it can be written in the form mk for some integer m. The greatest power of n is the largest number k for which n is a perfect power of order k. Some examples:

• 9 = 32 is a perfect square, and it cannot be written in the form mk for k > 2, so it's greatest power is 2.
• 16 = 42 is also a perfect square. However it can also be written as 16 = 24, so it's greatest power is 4 instead.
• 24 = 241 is not a perfect power of any order k > 1 so it's greatest power is 1.

The Challenge

Given a positive integer n > 1, determine its greatest power. This is OEIS entry A052409 (with a(1) defined as 0, but you don't need to handle that).

You may write a program or function, taking input via STDIN (or closest alternative), command-line argument or function argument and outputting the result via STDOUT (or closest alternative), function return value or function (out) parameter.

Standard rules apply.

The first 100 terms of the sequence (starting from n = 2) are:

1, 1, 2, 1, 1, 1, 3, 2, 1, 1, 1, 1, 1, 1, 4, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1,
3, 1, 1, 1, 1, 5, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 6, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 4, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1

• Can anyone find a similar challenge we've already done? – Martin Ender Feb 26 '16 at 18:08
• I thought up this exact challenge a few days ago, but didn't bother writing it out. Glad to see that someone else did :) – ETHproductions Feb 26 '16 at 18:13
• Power Points? :P – FryAmTheEggman Feb 26 '16 at 18:35
• Also I think I found it: codegolf.stackexchange.com/q/501/31625 although I would prefer to close that as a dupe of this one, I think. It has basically no spec... – FryAmTheEggman Feb 26 '16 at 18:37
• @FryAmTheEggman Hmmm, that might editable without invalidating answers though. – Martin Ender Feb 26 '16 at 18:47
• By the looks of it, some answers are already invalid, and a bunch give 0 for the values that, in yours, should give 1. Not sure what to do with that, I think it'd have to say that 1 or 0 was acceptable? – FryAmTheEggman Feb 26 '16 at 18:56
• @FryAmTheEggman I think even by the little bit of spec that's there, returning 0 for anything is wrong. Let's see if froddy responds to my comment. Otherwise, I'll probably go ahead with this one. – Martin Ender Feb 26 '16 at 18:57
• Rather than "greatest power" it's "greatest-order root". But of course that's not a catchy title :-) – Luis Mendo Feb 26 '16 at 19:23
• @DonMuesli I think it would be the greatest root-order, because the greatest-order root should be m, right? – Martin Ender Feb 26 '16 at 20:48
• @Martin Yes, you are right. I had understood the challenge the other way around. Very nice challenge BTW! (Meaning I think I have a 10-byte answer, hehehe) – Luis Mendo Feb 26 '16 at 21:34
• Submit to a greater power? – Digital Trauma Feb 27 '16 at 2:16
• There was a similar question about expressing your number n in all possible ways as n=a^b, but I'm not finding it now. – xnor Feb 27 '16 at 22:44
• @xnor Hmmm, let me know when you do. – Martin Ender Feb 27 '16 at 23:01
• @xnor There's codegolf.stackexchange.com/q/564/8478 also by froddy. – Martin Ender Feb 27 '16 at 23:02
• @MartinBüttner Here is is: codegolf.stackexchange.com/q/58047/20260 – xnor Feb 27 '16 at 23:08

**The Noether Challenge**

It is Emmy Noether's birthday. She was a pioneer in the field of ring theory.

The challenge is to compute two graph-theoretic invariants of a certain graph we can associate with any ring R.

For this challenge, we don't require a unity (multiplicative neutral element) in R, do require the commutativity of multiplication, and additionally require that the ring have finitely many elements. That is, we want to consider finite commutative rings, not necessarily unitary. From now on, "ring" will mean just that.

We will also need the concept of a zero divisor. A zero divisor of a ring R is an element r of R such that r*s=0 for some non-zero s in R. 0 is a trivial example of a zero divisor in any ring with more than one element since it can be shown that 0*s=0 always.

Something Emmy Noether didn't think about is zero-divisor graphs. They have, however, been quite extensively studied recently. The zero-divisor graph of a ring is simple, undirected graph formed as follows. The vertices are the the zero divisors excluding 0. Two vertices r and s are connected by an edge whenever r*s=0, excluding the cases in which r=s (that is excluding possible loops).

In this challenge you are given a file with (addition and) multiplication table(s) for some ring R as input. Your program in any language has to output the diameter and the girth of the zero-divisor graph of R.

The file's name is "ring" and it's a text file. You may assume any extension you wish. Depending on your preference, you may assume that the addition table is not present. If it is, it comes below the multiplication table. The ring is assumed to have at most 62 elements and the elements can be denoted by any subset of the alphanumeric characters including lower- and upper-case letters of the English alphabet. The only other characters assumed to be in the file are the whitespace, the newline, "+" and "*". The "0" character is reserved for 0, so you don't need to check which label stands for 0. The rows and columns for 0 come next after the label rows and columns in both tables. The character "1" doesn't have any special meaning.

The first row and the first column of each table are reserved for labels and the labels' order is the same everywhere (in both tables and both in the columns and in the rows). The upper-left-most character of the multiplication table is "*" and the upper-left-most character of the addition table is "+". The characters in either table are not separated. "*" is the first character in the file. You may assume anything you want about the numbers of whitespaces that follow each row of either table before there's a new line. If two tables are present, they are separated by exactly one additional newline. You may assume whatever you want about the number of newlines and whitespace after the last table.

The output is two numbers, in whatever human-readable and human-understandable form. And the form shouldn't make the user angry. They are to represent the girth and the diameter of the zero-divisor graph. If either of the invariants is infinite, again, it's up to you how you want to output them but it needs to be understandable. If you want to use a nonstandard symbol, like "i", for that, tell the user what it means. We assume that the input is valid. In particular, we assume that the operations in the file are actually ring operations! Your code should be ready to go in whatever way is standard for your language of choice. It shouldn't need any more code to run.

Reading up and research are encouraged, but if you want to use some non-obvious mathematical fact, please give a source or a proof. And in general, please explain how your program works. All standard loopholes are disallowed.

• The asterisk seems to make italics sometimes instead of just being text. Can you help me with this? Why does it happen and how do I escape it? – ymar Mar 23 '15 at 4:12
• Why it happens. The general solution is to escape with backslash. I think Martin's intention was that you mark up entire equations as code, not just the asterisk character. – Peter Taylor Mar 24 '15 at 10:24
• Comments on the question itself: 1. It's not obvious that this is a graph theoretic question rather than an algebraic question until a long way into it. Ideally the title should make that clear. 2. The first two ring axioms are clearly misstated. But 3. listing them seems to be a waste of space. The real question (given a multiplication table, identify the zero divisor graph and its diameter and girth) can be stated quite clearly without them. 4. Both diameter and girth can be infinite. How should that be represented in output? – Peter Taylor Mar 24 '15 at 10:41
• 5. A question shouldn't be popularity-contest unless there's no other reasonable way of scoring it. This would function perfectly well as code-golf, and with a minor change to allow arbitrary sizes of input could be a fastest-code. – Peter Taylor Mar 24 '15 at 10:42
• @PeterTaylor I've made corrections. – ymar Mar 8 '16 at 14:18
• Thanks. Looking back at this, I have two further comments: 6. It's generally considered best to be a bit more flexible on input mechanisms: i.e. rather than specifying that the input must be read from a file, specify its format as a string and leave the default formats open (so people can read from stdin, write a function which takes a string parameter, etc). 7. Some test cases would be good. – Peter Taylor Mar 8 '16 at 21:49

Igpā Atinlā (Pig Latin)

This is .

In this challenge, we will be translating strings of words to Pig Latin.

Input: A string of words (a "word" is a continuous sequence of the characters A-Za-z) (ASCII only).

Output: The translated version of the input. Translations described below.

Some Definitions:

First, the following are vowels: "A,a, E,e, I, i, O, o, U, u". Any alphabet character that is not a vowel is a consonant. A consonant cluster is any continuous sequence of consonants surrounded on both sides by non consonant characters (or beginning/end of input). An example, the clusters are in bold:

"I am two hundred years young, you child-mother."

A word is a set of alphabet characters surrounded on both sides by non-alphabet characters.

Translation:

1. If the string "'s" or "'d" or "'t" appears ("apostrophe s/d/t"), remove the apostrophe.

For each word in the string, do the following:

1. If a word only contains capital letters (A-Z), ignore the next step.

2. If the beginning of the word was a capital letter (A-Z) AND a consonant, change it to its lower-case equivalent (a-z). Then capitalize the first vowel in the word. If no vowel exists, then recapitalize the letter. e.g. Stretch --> strEtch, "Twxx" --> "Twxx"

3. If the word begins with a consonant cluster, move that consonant cluster to the end of the word. e.g. stretch --> etchstr

4. Append the "hard a" character to the word. If every letter is capital, append 'Ā'. Otherwise append 'ā' e.g. etchstr --> etchstrā, "ATC" --> "ATCĀ"

• If your language is unable to output 'ā' and 'Ā', you may use "ay" and "AY" respectively.
5. If the word is "A" or "a", ignore previous instructions. Transform the word to "Anā" or "anā", respectively. (because in Pig Latin, everything begins with a vowel, so we use the article "an" instead of "a")

Test Cases: (I think these are right)

• "I want to be a cat." --> "IĀ antwā otā ebā anā atcā"

• "That's a really nice... ass-car?" --> "Atsthā anā eallyrā icenā... assā-arcā?"

• "[CR][NL]'ssssssssssssssssssTRUExxxxxxxxIAMSOCOOL" --> "[CRĀ][NLĀ]UExxxxxxxxIAMSOCOOLssssssssssssssssssTRā"

• "THIS CHALLENGE IS PROBABLY A DUPE" --> "ISTHĀ ALLENGECHĀ ISĀ OBABLYPRĀ Anā UPEDĀ"

• "I SAT ON an APPLE" --> "IĀ ATSĀ ONĀ anā APPLEĀ"

• Is the string: bc a consonant cluster? – Downgoat Mar 6 '16 at 5:30
• Yeah. "bc" --> "bcā". Is that not what the spec says? – Liam Mar 6 '16 at 5:32
• ok, I wanted to clarify if the start/end of a string counts as a "non-vowel" character. Perhaps a better description is: "a sequence of consonants"? – Downgoat Mar 6 '16 at 5:34
• I think it's clearer now? – Liam Mar 6 '16 at 5:52
• – FryAmTheEggman Mar 7 '16 at 15:46
• @FryAmTheEggman I think mine is sufficiently more complicated that it wouldn't be a dupe. However I don't know that the complexities are interesting enough to merit posting. – Liam Mar 7 '16 at 19:22
• Some languages might not support the special a and A. Maybe allow regular A's? – CalculatorFeline Mar 9 '16 at 5:43
• @CatsAreFluffy I know that some languages can't but I think that makes the challenge more interesting. At least a bit. If people disagree, I might change it to "ay" and "AY" – Liam Mar 9 '16 at 5:49
• In the THIS CHALLENGE IS PROBABLY A DUPE test case, the A should become Anā instead of ANĀ. – user48538 Mar 9 '16 at 7:07
• I'd call this a dup. – mbomb007 Mar 11 '16 at 15:29

Maximum of Two Roman Numerals

You should write a program or function which returns the maximum of two Roman numerals.

Input

• Two positive integers between 1 and 3999 (inclusive) with their Roman numeral representation string.
• The two strings can be separated by a space or inputted in the standard list representation of your language.
• 4, 9, 40, ... are written as IV, IX, XL, ...
• Trailing newline is optional.

Output

• The larger Roman numeral as string.
• Trailing newline is optional.
• If the two inputs are equal you should still only return one of them.

Examples

Format is input => output (explanation)

XXIX DI => DI (29 < 501)
V X => X

TODO more


Built-in functions involving Roman numerals are prohibited.

This is code golf so the shortest entry wins.

• Can I adopt this abandoned challenge? – programmer5000 Jun 9 '17 at 12:12
• @programmer5000 Yep! – randomra Jun 10 '17 at 17:02

Entropy Golf

This is a scoring system without a challenge.

My idea is to score entries based on the total Shannon entropy contained within them. This provides an incentive to both use fewer unique characters and to have a shorter program overall.

Given a string of characters, the score is calculated as follows, where C(x) is the number of occurrences of the given letter. To help provide a correction for multi-file programs or languages in which the program length encodes information, the EOF character at the end of every file is to be counted for the purposes of this scoring mechanism. Lowest score wins.

$$\mathrm{score} = -\sum_x{C(x)\log_2\frac {C(x)}{\mathrm{Length}}}$$

Anybody who knows better notation/MathJax is free to edit.

• As it is you would still get many Lenguage-like answers in the form of unary string, base convert, exec. Edit: Actually that wouldn't work, never mind. – feersum Jun 6 '15 at 15:51
• The line between lenguage and slashes and brainfuck and real languages would be hard to define. Languages with fewer symbols tend to require more uses of those symbols to accomplish anything. – Sparr Jun 6 '15 at 17:01
• @Sparr It's mainly about languages that encode information in the length of the program itself. It's not about number of symbols, although I could probably put a requirement that the program uses at least two symbols. – PhiNotPi Jun 6 '15 at 18:05
• Alternatively you could consider each program to be terminated by an EOF character, which is included in the calculations. Then a $n$-character program in lenguage would score $-n \lg n/(n+1) - \lg 1/(n+1)$, which does grow indefinitely. PS I've just realised that my edits come down to the same thing as you were doing by flipping the quotient in the log upside down. Sorry about that. Revert if you want, but be aware that I might not be the only person who has $p \lg p$ so branded in their mind that anything else looks wrong. – Peter Taylor Jun 6 '15 at 18:46
• @PeterTaylor p log (1/p) should be much more intuitive for people who see this the first time. – jimmy23013 Jun 7 '15 at 1:20
• If you meant unique characters, Sclipting has an advantage for that selecting a character from the whole Unicode contains a lot of information (compared to the length of a short program). – jimmy23013 Jun 7 '15 at 1:23